Economic recovery from the Argentine great depression: institutions, expectations, and the change of macroeconomic regime

•

•

,... .,..FUNDAÇÁO ,... GETULIO VARGAS EPGE

Escola de Pós-Graduação em Economia

"Economic Recovery from the Argentine

Great Depression:Institutions,

Expectations, and the change of

Macroeconomic Regime .. "

Prof. Alan Taylor (Northwestern University/NBER)

(co-autoria com

Gerardo della paolera)

LOCAL

Fundação Getulio Vargas

Praia de Botafogo, 190 - 10° andar - Auditório

DATA

13111198 (6

_feira)

HORÁRIO

11 :OOh

. ,r.

I

/

I .

Economic Recovery from the Argentine Great Depression:

Institutions, Expectations, and the Change of Macroeconomic Regime'

Abstract

First Draft, May 1998

Gerardo della Paolera, Universidad Torcuato Di Tella and Hoover Institution t

Alan M. Taylor, Northwestern University, Hoover Institution, and NBER"

This work explores how Argentina overcame the Great Depression and asks whether active

macroeconomic interventions made any contribution to the recovery. In particular, we study Argentine macroeconomic policy as it deviated from gold-standard orthodoxy after the final suspension of convertibility in 1929. As elsewhere, fiscal policy in Argentina was conservative, and had little power to smooth output. Monetary policy became heterodox after 1929. The first and most important stage of institutional change took place with the switch from a metallic monetary regime to a fiduciary regime in 1931; the Caja de Conversión (Conversion Office, a currency board) began rediscounting as a means to sterilize gold outflows and avoid deflationary pressures, thus breaking from orthodox "mIes of the game." However, the actual injections of liquidity were small' and were not enough to fully offset the incipient monetary contractions: the "Keynes" effect was weak or negative. Rather, recovery derived from changes in beliefs and expectations surrounding the shift in the monetary and

･ＺＭＺ￧ｾＺｭｧＺＬＭＡＢ｡ｴ･＠ regifl1e, anel the deli'1king of gold flows and the money base. Agents perceivod 2. ne'."

regime, as shown by the path of consumption, investment, and estimated ex ante real interest rates: the "Mundell" effect was dominant. Notably, this change of regime predated a later, and supposedly more significant, stage of institutional reform, namely the creation of the central bank in 1935. Still, the extent of intervention was weak, and insufficient to fully offset externaI shocks to prices and money. Argentine macropolicy was heterodox in terms of the change of regime, but still conservative in terms of the tentative scope of the measures taken .

• This paper is part of a larger project on Argentine interwar macroeconomic history supported by a grant from the National Science Foundation administered by the National Bureau of Economic Research. This paper was prepared whilst Taylor was a National Fellow and della Paolera was a Visiting Fellow at the Hoover Institution. We gratefully acknowledge financiai support from the National Science Foundation, Hoover Institution, and the Centro de Investigación en Finanzas y Mercado de Capitales (CIF) at the Universidad Torcuato Di Tella. For their research assistance we thank Sandra Amuso and Marcela Harriague.

-- t Gerardo della Paolera is Rector and Professor of Economics at Universidad Torcuato Di Tella, and a Visiting

Fellow at the Hoover Institution, Stanford University. Address: Minones 2159/77, Capital Federal, Buenos Aires 1428, Argentina. Telephone: 784-3386. Fax: 784-0089. Email: gpaolera@mail.retina.ar.

• tt Alan M. Taylor is an Assistant Professor at Northwestern University, a 1997-98 National Fellow at the

Hoover Institution, Stanford University, and a Faculty Research Fellow at the National Bureau of Economic Research. Address: Department of Economics, Northwestern University, 2003 Sheridan Road, Evanston, IL 60208-2600, United States. Telephone: 847-491-8234. fax: 847-491-7001. Email: amt@nwu.edu .

, '

. ---- -

.-{tI.UOTECA MARIO.

tiEMftlQUE

ｉｉｉｾ＠

•

Introduction

The experience of Argentina during the Great Depression provides an ideal historical laboratory for the investigation of macroeconomic stability and policy choice in a small open economy under a fixed exchange rate regime. Given recent developing country experiences in Latin America and Asia, inciuding examples of contemporary monetary regimes with currency boards very similar to

Argentina's institutions of the early twentieth century, there is considerable topical interest in these questions. The essential questions is: what happens if you employ a currency board and there is an externai crash or deflation threat? This was the nature of the crisis in the 1930s for many countries, and the same potential problem faces Argentina, Hong Kong, and other countries in the 1990s. What should they do today? To inform that question we ask: What did Argentina do in the past?

The Great Depression began in Argentina in the late 1920s, even before the traditional date for the onset of the Depression in the core, the Wall Street crash of 1929. Like many countries of the periphery, Argentina was exposed to the commodity lottery. As Díaz Alejandro (1983) and

Kindleberger (1986) have noted, this exposure led to macroeconomic fortunes collapsing as the terms of trade worsened through the 1920s. By December 1929, the balance of payments crisis was severe, and the exchange rate was allowed to float after a mere two-year resumption of the gOla standard. But this was not a decisive break from macroeconomic orthodoxy. Fiscal policies remained

conservative under Uriburu, and, even more than in other countries like the United States; we even find evidence of fiscal tightening just as the worst of the Depression hit during 1929-31 (Ortíz 1993).

Recovery began in 1931, as output grew for the first time in several years, and by 1934-35, output had regained its 1929 leveI. We show that assigning fiscal policy any responsibility for this recovery is implausible: by any measure fiscal policy actually tightened during the early 1930s, as in many other Latin American countries, and as in the United States. Orthodoxy in fiscal policy was not an immediate casualty of the change in regime. Thus we move on to examine monetary policy actions from 1929 to 1935, seeking for evidence of a change of regime. Although many commentators would see the creation of the Banco Central (Central Bank) in 1935, according to Federico Pinedo' s plan, as the main monetary policy event of the 1930s in Argentina, we instead emphasize the remarkable decision of the Caja de Conversión (Conversion Office, a currency board) to began rediscounting and so forge an independent monetary policy, as early as 1931, at the urging of Raúl Prebisch. This decision to sterilize gold losses to offset monetary contractions was the decisive break with the old orthodoxy. In many ways, the later creation of the Central Bank merely rubber-stamped the

operations of this new macroeconomic policy regime, and continued these operations after 1935. I But

the Argentine recovery was complete by 1935; and the only pre-1935 change in regime that could be assigned a role in ending the Argentine Great Depression was the change in ideas at the Conversion Office. Yet, did it make a difference?

To investigate the question, we conduct an econometric analysis of prices, exchange rates, and interest rates. We find that the change of monetary regime was essential to Argentina's recovery in that it helped avert a devastating collapse of prices, and, potentially, of output in 1931-33. Instead of following the United States and other countries into this abyss, Argentina' s regime shift destroyed deflationary expectations, pennanently lowering previously extremely high real interest rates. In other ways, though, policy was stilllimited by orthodox thinking. Sterilizations offset gold outflows to a large degree, but never counteracted them to any great degree: indeed the money base barely changed from 1929 to 1935, though at least it didn't shrink as much as orthodoxy would have required. So there was no "Keynes" effect at work, no large money injection to stimulate aggregate demand via the money market. In this sense, Argentina was still a prisoner of its intellectual and economic history, and the Conversion Office, though willing to follow Prebisch' s plan, was not willing to pus h it

as far as it might to use monetary expansion as a de vice to end the Depression more quickly. U sing monetary policy, it was apparently very hard to break from the constraint of purchasing power parity (PPP) at this stage, so aggregate demand was hard to manipulate.

In the end, the key channel through which the change in monetary regime had real effects was via the destruction of deflationary expectations. With nominal interest rates subject to the zero floor, this change of regime could significantly reduce ex ante and ex post real interest rates. This was indeed the case, promoting recovery through increases in aggregate demand via investment and consumption activity: the "Mundell" effect of the change in expectations was the vital source of recovery. Hence we think that the institutional change heralded by the rejection of an old orthodoxy was just as essential to recovery from the Great Depression in the periphery as in the core.

Contours of the Argentine Great Depression

The Great Depression marked the end of an epoch where free trade and integration into externai capital markets acted as the main recipes to secure economic growth and prosperity for the Latin American Countries (Díaz Alejando 1983, 1984). In important studies, O'Connell (1984) and De Paiva Abreu (1984) noted how the dramatic change in the international economy shaped political-economy views about the welfare enhancing aspects of an outward economic orientation in two of the most important Latin American countries, Brazil and Argentina. In the deteriorating cIimate for trade

I In substantive terms, we could argue, the main contribution of the central bank was to put in place a rescue

package for the financiai sector using the proceeds of gold revaluation.

3

,

,

in the 1930s, both countries tried to structure bilateral trade agreements with the two world powers, United Kingdom and the United States, with mixed results. The asymmetric bargaining power of the new international economic order took its toll on their domestic economies. For Argentina, although economic problems had retarded economic growth since 1914 (Taylor 1992), there was a pressure to stand by the old liberal orthodoxy. Even during and after the worst recession in Argentine history, during World War One, few policymakers seriously questioned the return to orthodoxy as the goal for the 1920s, to rebuild an economic order predicated on openness in markets and adherence to the gold standard as a monetary rule.2 _{This view was only broken as the threat of an even worse economic}

collapse loomed in the early 1930s.

In Table 1, we depict some important macroeconomic data that characterizes the impact of the Great Depression on the Argentine economy. From 1929 to 1932, Argentina imported the severe deflationary pressures in the international economy. 3 The externai terms of trade declined by 24% and

the trade share, defined as the ratio of the semi-sum of exports and imports to real output went from an ali-time high of 36% in the late twenties to a low of 28% in 1932. If one were to judge by its trade

exposure, Argentina was one of the most vulnerable economies in the presence of such sizeable foreign shocks-shocks that, in addition to pure deflation, also included fierce terms of trade declines as countries took their hits in the "commodity ｬｯｴｴ･ｲｹＮＢｾ＠

However, it is astonishing is that the Argentine Great Depression was so mild and short-lived by international standards. Hence, the notion that an important change in economic policy took place, and saved Argentina from more pronounced suffering, deserves close scrutiny. As can be gleaned fram Table 1, from peak to trough (1929 to 1932), the domestic real output fell by "only" 14% and had even surpassed its 1929 levei already by 1935. Deflation, a curse to avoid in the inter-war period, was only about 6% in the 1929-1932 period. The behavior of output and prices compares favorably, say, to North American gold-standard countries such as the United States and Canada: they had an overall decline in real activity of more than 30% fram peak to trough, and a decline in their price levei of more than 20%.5

2 Though rarely remarked on, this point about the relative size of Argentine recessions bears repeating. We can

measure the size of recession by a simple technique, cumulating deviations below the previous peak levei of log output over subsequent years, until the previous peak is surpassed, and measuring the cumulative loss as a fraction of annual output. Using this yardstick, the three largest recessions of the period from 1884 (the start of annual output data) to 1940 are the Baring Crisis of the years 1890-93 (4 years, cumulative loss 31 %), the recession of 1914-19 (6 years, 63%), and the Great Depression of 1930-34 (5 years, 43%). See della Paolera and Taylor (1998).

3 On these deflations, see Kindleberger (1986); Temin (1989); Eichengreen (l992a).

ｾ＠ On the commodity lottery see Díaz Alejandro (1984). For a discussion of the general experience of the

periphery in the 1920s with terms-of-trade shocks see Kindleberger (1986).

5 The data are fram Mitchell (1992; 1993).

4

Table 1

Contours of the Argentine Great Depression

(a) Nominal Variables

Money Base Money General Land Banks'

Total Gold Domestic Supply Exchange Price Wage Price Discount

• (MO) Stock Credit (M3 ) Rate Levei LeveI LeveI Rate

m $mn m$mn m$mn m $mn $mn/$US 1913=100 1913=100 1913=100 o/c

1913 823 530 293 1,687 2.35 100 100 100 5.4

1928 1,406 1,113 293 4,717 2.32 131 180 296 6.3

1929 1,247 954 293 4,652 2.35 127 178 293 6.9

1930 1,261 968 293 4,660 2.70 122 166 238 6.9

1931 1,245 593 652 4,149 3.40 118 155 260 7.2

1932 1,339 584 755 4,116 3.83 119 146 237 7.1

1933 1,214 561 653 4,061 3.18 114 139 210 6.1

1934 1,172 561 610 4,078 3.89 130 136 196 5.5

1935 1,647 1,354* 293t 4,180 3.75 128 147 185 5.4

1936 1,685 1,528* 157t 4,611 3.55 131 153 217 5.6

1937 1,679 1,422* 257t 4,922 3.28 150 159 252 5.2

1938 1,615 1,296* 319t 4,811 3.86 140 160 271 5.3

1939 1,796 1,396* 400t 4,960 4.27 143 166 256 5.8

1940 1,810 1,329* 481t 5,050 4.30 163 176 248 5.8

(b) Real Variables

Real Components of GNP Consolidated

Terms Exchange millions of $mn 1913 Government

ofTrade Rate Output Privo Con. Gov. Con. Investmt. Exports Imports DeficitlGDP

1913=100 1913=100 Q C G I X M %

1913 100 100 4,640 4,322 204 579 1,805 2,270 0.8

1928 99 121 7,780 6,549 406 900 2,901 2,991 1.7

1929 90 126 8,146 6,781 425 1,029 2,847 3,048 2.3

1930 88 144 7,784 6,829 408 871 2,100 2,533 4.3

1931 65 165 7,216 5,248 393 533 2,871 1,651 2.7

1932 68 162 6,966 5,092 393 374 2,636 1,282 l.8

1933 64 139 7,309 5,723 415 418 2,474 1,506 1.7

1934 79 159 7,912 6,085 447 554 2,546 1,584 l.8

1935 79 153 8,275 6,187 538 691 2,754 1,836 l.2

1936 96 148 8,336 6,249 565 824 2,491 1,870 1.7

1937 110 121 8,964 7,145 612 748 2,911 2,381 2.4

1938 101 151 8,979 7,703 634 824 1,963 2,251 2.4

1939 89 162 9,337 7,271 668 691 2,501 1,755 3.6

1940 91 145 9,486 7,588 672 637 2,071 1,461 2.7

t Domestic Credit is defined as after 1935 as money base minus revalued gold minus reserves.

* After 1935 gold reserves were revalued when official par changed from 2.27 to 4.96 paper pesos per gold peso.

Source: della Paolera and Taylor (1998).

5

•

't

The Argentine performance is also very good by the standards of the peripheryo Other Latin American countries were much more affected by the downturno In the same 1929-32 period, Mexico's prices and output fell by 199'0, Chile's real output by 279'0, and Brazil's by 289'00 Here it is interesting to note that the mere presence of currency depreciations of important magnitude is not directly related to the extent of the experienced depressiono For example, in Brazil there was a 669'0 depreciation of its currency, in Mexico 479'0, while the Argentine paper peso declined by 639'0 with respect to the gold dollaro

Thus, it is important here to distinguish between those currency depreciations that were part of a change in the macroeconomic regime and others that merely reflected typical foreign exchange market pressure in an inconvertible regime during a severe downturn in the balance of paymentso As we shall see later for the Argentine case, there is an enormous difference between: (a) just letting the currency depreciate as a mean to restore equilibrium in the money market; and (b) installing a new policy regime à la Sargent (1983) that will influence expectations and hence alter the course of

economic decisionso The distinction is between a country choosing to depreciate as part of a regime switch versus being forced to depreciate whilst adhering to orthodoxyo This was recognized by della Paolera and Ortiz (1995): " .. o the 1931 year, however, was without doubt a peculiar one: the year started with the combination of an extremely orthodox domestic fiscal policy and an unequivocal convergence towards a fiduciary monetary system .. 0"0 In other words, if she were free to choose to effect a full-fledged change in her monetary regime, as a proactive polítical economy decision, Argentina must have had a lot of room for maneuver. Indeed she did: years of orthodox adherence to the gold standard had led to a massive build up of gold reserves, and an enormous backing for the domestic currencyo

From Table I we note that around 1930, almost 80% of the money base was backed with goldo The inelastic relationship between gold and money finally broke in 1931 when the government decided to switch from targeting the gold reserves backing the quantity of money to targeting the nominal quantity of money itselfo In this manner, the Argentine ｡ｵｴｨｯｲｩｴｩｾｳ＠ halted <: decline in the suantity of money in a bid to avert deflationo The gold stock was criticaI. The use of gold for fiscal purposes, to service externai debt obligations, allowed the government to maintain a very orthodox fiscal policy by using these tied-up resources; thus did Argentina escape default on foreign debts in the 1930s, a very unusual feat for a peripheral country o So the new policy mix was chosen: sound finance in the realm of fiscal affairs, and, at the same time, an unorthodox fiduciary monetary regimeo In 1931 and 1932, the domestic credit component of the money base, frozen at 293 million pesos for 32 years, increased to such an extent that it already accounted for 62% of the money base by 19320

Other macroeconomic responses stand out in Table lo Investment fell by about one half in 1929-31, then by another third in 19320 Private consumption fell by about a quarter. Both shifts greatly exceeded the fall in outputo One can understand the perceived need to implement some change

6

,

in policies to redress the negative expectations that prevailed. The banks' interest rate was pretty flat at around 7%, so an injection of liquidity could perhaps ease the credit market via the "Keynes effect." But most of the effort to change expectations had to come through the "Mundell effect"-the destruction of deflationary expectations. fi

For Argentina, it might be observed that the dramatic declines in output, investment, and consumption ceased, and then reversed, after 1932 (Table 1). The evolution of the real exchange rate suggests that the authorities were successful in the regime change, and were able to anchor the real exchange rate in spite of the sizeable deterioration in the terms of trade. Export growth and sustained import compression assisted recovery. In short, by 1933 Argentina had circumvented the most

devastating effects of the World Depression.7 _{Did policy choices make a difference? In the next}

section we examine the features of Argentine fiscal and monetary policies.

The Interwar Gold Standard: Orthodoxy and Heterodoxy

In important studies, Díaz Alejandro (1983) and Ortiz (1993) have characterized the Argentine Great Depression as a definite "blessing"-in the sense of creating an opportunity to adopt new economic policies and institutions. This new institutional regime, by deviating from the prevailing mentalité of the orthodox gOla standard, could insulate the domestic economy from the dlsmal global ｳ｣･ｮ｡ｮｯＮｾ＠

Such a reaction was hard to envisage just a few years earlier. In this section we discuss Argentina's position as a small open economy at the periphery and her struggle, over more than a century, to securely establish credible monetary, fiscal, and financiai institutions. This effort followed from a political consensus in which few seriously doubted the rewards that would (and did) accrue to Argentina in return for embracing globalization in the late nineteenth century and playing by the "rules of the game" as laid down by private finance and govemments in the core: free movements of capital, labor, and goods were a key ingredient in Argentina' s pre-1914 succ.:ess. Against this

backdrop, and despite the instability in the global economy after 1914, it should still be apparent what extraordinary intellectual and technocratic obstacles policymakers had to surmount in the 1930s before they could effect a radical change in the macroeconomic regime, and depart from the prevailing orthodoxy even ahead of most of the developed countries of the core. In this context, the timing and

fi Temin (1989) noted the impact of these two effects in the 1930s recovery in the core countries. and he toa

stressed the importance of the Mundell effect and the change of expectations that followed the change of regime (see a1so Eichengreen 1992a; 1992b). The U.S. experience under Hoover and then FDR is a c1assic example of such a regime shift (Temin and Wigmore 1990; also Romer 1992).

7 See however the impact on the economy of weak financiai institutions in della Paolera and Taylor (1997)

8 Diaz Alejandro (1983) states: " ... Once upon a time foreign money doctors roamed Latin America prescribing

fixed exchange rates and passive gold exchange standard monetary rules. Bankers followed in their footsteps, from the halls of Montezuma to the shores of Daiquiri ... This paper will chronic1e some of the ways various Latin American economies coped with them. It will be seen that the performance of several [Latin AmericanJ economies was remarkably good , under the circumstances ... "

•

,

depth of institutional changes are crucial to determine when and how policy actions modified behavior and expectations in the domestic economy, and we therefore spend some time surveying the evolution of fiscal and monetary policies.

Fiscal Policy

Whereas the above discussion highlights the changing context of monetary policy over the 1930s, the historical record shows no similar evolution towards an interventionist stance in the sphere of fiscal policy. As we shall show presently, Argentina' s basic orthodox fiscal stance, of seeking to maintain budget balance, was conserved even during the worst years of the Great Depression. For a number of reasons we should not consider this outcome surprising.

First, even in the economies of the core, the power of fiscal policy was not fully unleashed to insulate economies from the recession. Thus, for example, the classic Keynesian tool of

macroeconomic management did not fail, but rather was never really used, as was famously pointed out by Brown (1956).Y Simply put, for all the fanfare about the impact of New Deal spending

programs, it was hard to find any evidence of full employment deficits in the U.S. in the 1930s. Indeed, for many years, the net fiscal impact appeared to be even slightly contractionary.

Second, even if fiscal activisrn was ar option for core c:ountries in thp 1 _{Q1{)c ",ith} their

developed taxation systems and fiscal sophistication, it was likely to be much less of a feasible policy choice for countries at the periphery. Peripheral countries simply had much less developed

govemment structures for managing, administering, and implementing large spending programs. This was the case throughout Latin America, for example, where Twomey (1983) discovered that few countries in the 1930s were capable of developing new fiscal programs in response to the Great Depression.

There is no strong prior reason to expect Argentina to deviate from this more general pattem. Argentina had kept up a much more rigid historical adherence to orthodoxy and the "rules of the game" under the classical gold standard. This was in an attempt to win credibility from core countries, gain access to foreign capital (successfully), and pursue a greater degree of macroeconomic stability than other countries in the region. Thus, there is, arguably, every reason to expect Argentina to have been even more fiscally orthodox than her neighbors up to and even during the 1930s. The 1930s fiscal experience in Argentina more or less accords with expectations:

Tax revenues lagged behind expenditures during President Yrigoyen's administration; in 1930 nominal tax revenues, heavily dependent on import duties, fell in absolute amounts .... Large deficits were registered in 1930 and 1931, which could be regarded as being induced by the decline in foreign trade rather than as autonomous acts of policy .... As in other Latin American countries fiscal heterodoxy was discredited in Argentina by lax budgets during the late 1920s.

9 This finding was further reinforced by the work of Peppers (1973) and Romer (1992).

8

'. t.,...

• : J ｾ＠

,

Figure 1

Argentine Fiscal Policy

1.600 ｾ＠ _ .. _ • Taxes-Customs

••.... Taxes-Income and Wealth ---Taxes-Total

1.400 .

-+-Expenditure-Total

1.200

1.000

800

600

-,

ＭＭＮ｟ＭＭＭＮｾ＠ ...

o -' .

..

. ----.-.

.---

'."''-

. -.-.-...

'.' . '' ...: " .

-1913 1917 1921

Notes: Units are millions of paper pesos. Source: Mitchell (1993).

1925 1929 1933

...

1937

Both the Uriburu and the Justo administrations attempted to reduce expenditures and to increase taxes during the early 1930s; an income tax was introduced in 1932 and tariff rates were increased earlier. (Díaz Alejandro 1983, 21)

Various data sources confirm this view. One simple measure is the fiscal action of the central

government, shown in Figure 1. It is apparent that at the start of the interwar period customs taxes

levied on imports constituted a large share of revenues, as is typical in ali developing countries,

accounting for $199 million out of $370 million pesos of tax revenues in 1913, or about 54%.

Consequently, tax revenues were likely to be cyclically correlated with trade conditions, and thus

general economic conditions, since during recessions, import contraction was a standard response.

This is borne out by the data, with dramatic declines in customs taxes during the WWI crisis, in the

1921-22 and 1926 recessions, and, strikingly, after 1929 and throughout the 1930s. As govemment

spending increased more or less on a trend in this period, there was thus some scope for an

endogenous (though not, by definition afull employment) deficit to appear in these recessions, and

indeed temporary deficits were run up in W orld War One and in 1921-22.

9

•

,

Figure 2

Deficits as a Share of GDP

0.05 ｾ＠ _{--Deficit/GDP}

0.0'+ ｾ＠

0.03 ｾ＠

0.02 ｾ＠

0.01 ｾ＠

o

-0.01 ｾ＠

(\ "2

1900 1910 1920 1930 1940

Source: della Paolera and Taylor (1998).

The fiscal response in the 1930s was not so forgiving, as is also shown by Figure 1. High expenditures were run up in 1928-30, but were drastically cut back during 1931-33, generating a big contractionary effect just as the economy fell into the Great Oepression. There was no Argentine New Oeal as far as central govemment program expenditure was concemed. At the same time, althuugh I.:ustoms ldxes were faliillg in line -vith the trade crisis, total taxes were increasing.

President Uriburu, like Hoover in the United States, was a fiscal conservative, and was searching for a means to close the fiscal gap, and retum closer to an orthodox budget balance. One important part of the package was a dramatic increase in direct taxes, in the form of income and wealth taxes, which rose from a typicallevel of $25 million pesos (less than 5% of ali revenues) in the 1920s, to $92 million (almost 15%) in 1933. With a broad array of aggressive tax programs, the govemment raised taxes consistently every year after 1930, and closed the deficit from $240 million pesos in 1929, to just $126 million in 1933.

These conclusions hold up when an appropriately normalized measure of fiscal stance is calculated: we use the ratio of the deficit to GOP. Figure 2 shows this data using information now on the overall consolidated deficits of the entire public sector going back to 1900. Comparing the classical

10

..

,

gold standard years with the interwar and even the interventionist years of the 1930s reveals no marked shift in the propensity of the Argentine government to employ deficit finance, or apply fiscal policy as a countercyclical measure. And again, it is apparent that after 1930, fiscal policy so measured became ever more contractionary in both actual, and, thus, even more so, in adjusted full-employment terms .

Far from pursuing expansionary fiscal policy via increased fuI! employment deficits, the Argentine fiscal response during the Great Depression was such as to generate not even increases in

actual deficits, but rather a move toward surplus. We concur with Díaz Alejandro (1983, 22), that

"[i]n short, there is no evidence that during the early 1930s the Argentine government sought to increase the full employment budget deficit as a means to compensate for the fall in aggregate

demand." The net effect of fiscal actions were, then, surely contractionary, and remained so through at least 1935. The empirical search for a source of Argentine economic recovery must focus elsewhere, therefore, and as our earlier evidence suggests, must concern itself with the escape from gold-standard orthodoxy embodied in the change in monetary regime.

In fact, fiscal orthodoxy and monetary tensions were two sides of the same coin. Intense fiscal pressure was felt in many Latin American countries in 1929-30. Tax revenues fell and foreign lenders refused to rall over debts in the worsening international climate. Deflations and depreciations raised the burden of foreign-denominated debt service. The deficit for 1930 stood at 4.3% of output, a marked increase from previous years (Table I). Debt service had risen from 18 % of the budget in 1930 to 29% in 1932 Díaz Alejandra (1983, 20-21). Many countries chose to default in this situation, but

Argentina neve r wavered from her externai sovereign debt obligations, and here fiscal policy and commercial policy were ais o tied together. With so much of her trade linked to Britain, and with Britain imposing imperial trade preferences under the Ottawa accord, Argentina was left to plead for some trade concessions from a weak position. She finally obtained unfavorable terms under a mostly bilateral trade deal, the Roca-Runciman pact of 1933 (Díaz Alejandro 1983,28-29; Salera 1941). So :::; :-:ot to derail this crucial trading pact, Argentina maintained debt service, undcst:!nding th:lt British creditors could not be let down or else severe trade penalties might result fram a diplomatic crisis .

.The fiscal gap could not be closed by default, and attempts to raise tax revenues proved insufficient. The only remaining answer was to activate the last fiscal resource left to the government: the large gold stock sitting idle in the vaults of the Conversion Office. This course was taken, but the price for subscribing to this fiscally orthodox response was to draw down to gold stock and, ipso

facto, the money base, given the mechanical rules of operation followed at the Conversion Office.

Thus a deflationary money contraction was an inevitable but undesirable side-effect of the fiscal course chosen-at least so long as the Conversion Office played by its own rules.

li

•

Table 2

Monetary Policy Chronology, 1883-1935

1883

1883-1885

1886-1891

1891-1899

1899

1899-1914

1914-1927

1927-1929

1929-1931

1931

1935

Manetary Palie)'

Law of 1881 implemented, establishing gold standard.

Gold standard; par is 1 gold peso = 1 paper peso.

Baring Crisis; inflation leads to collapse of convertibility; exchange rate

begins to float and depreciates markedly.

Inconvertible paper currency. Floating exchange rate of paper to gold pesos.

Convertibility law. Creation of Caja de Conversión (Conversion Office), a

currency board. Currency board exchanges gold for pesos at new par (2.27) for ali transactions.

Gold standard; 1 gold peso

=

Inconvertible paper currency. Floating exchange rate of paper to gold pesos. Gold flows at the Conversion Office limited to occasional government uses. Currency board continues to exchange gold for pesos at new par for these transactions.

Gold standard; 1 gold peso = 2.27 paper peso.

Inconvertible paper currency. Floating exchange rate of paper to gold pesos. Gold flows limited to fiscal uses (payment of government foreign debt). Currency board continues to exchange gold for pesos at new par (2.27) for these transactions.

Currency board devi ates fram mechanical money creation rule: starts to l!se rediscounts. Start of independent Argentine monetary policy.

Creation of Banco Central (Central Bank). Takes over ali assets and

liabilities of the Conversion Office. Revalues gold stock according to prevailing market rate of exchange (new par is 4.96 versus 2.27). Uses proceeds to increase backing of money base, and to bail out financiai system.

Historically, from 1899 until the founding of the Central Bank in 1935, Argentina was under a currency-board regime (Table 2). The Caja de Conversión (Conversion Office) was the institution responsible for the administration of the domestic monetary regime. The Conversion Office held the typical macroeconomic responsibility of a currency-board: during periods of convertibility, they followed the orthodox gold-standard rules of the game, standing ready to automatically exchange domes ti c paper money for specie at the specified fixed nominal value of the domestic currency in terms of gold. In other words, the Conversion Office was just that, a window that mechanically and

instantaneously swapped paper pesos for specie. By adopting the gold-standard system, the monetary authorities unilaterally anchored the externaI value of the currency. Hence, they could not control the nominal quantity of money in the economy, and this quantity became an endogenous variable that depended on domestic and world money market conditions.

12

-•

Table 3

Changes in Gold Stocks and the Money Base, 1900-1935

AlI months Months with DG>O Months with DG<O Means

Regime Dates b SE N b SE N b SE N DG DM

Gold Std. 1900:2-1914:7 1.00 0.01 174 0.99 0.02 87 0.99 0.02 52 2.56 2.57

Float 1914:8-1919: 12 1.00 0.01 65 1.00 0.01 37 -0.58 0.83 5 6.74 6.74

Float 1920:1-1927:11 1.00 0.06 9S 1.00 0.04 15 l.30 0.39 8 1.90 1.90

Gold Std. 1927:12-1929:12 1.00 0.01 25 1.03 0.03 9 1.00 0.01 16 -4.43 -4.43

Float 1930: 1-1931:3 1.00 0.00 15 1.00 0.00 2 1.00 0.00 12 -4.42 -4.42

Float 1931:4-1935:4 0.05 0.24 49 0.47 0.43 13 -6.66 0.57

Note: Units of G (gold stock) and M (money base) are millions of paper pesos, with G evaluated at parity of 2.27 peso papel per gold peso. Regression is of DM on DG, using OLS, and reporting coefficient b, standard error SE, and sample size N.

Source: della Paolera and Taylor (1998).

It is interesting to see how c10sely the Conversion Office adhered to these rules, and we can use high-frequency monthly data to probe the questiono In apure gold standard regime, expansions and contractions of the nominal quantity of money are exactly correlated with variations in the gold stock at the Conversion Office. For the 1900-1914 and the 1927-1929 gold standard periods, there is a one-to-one robust association between money and gold as shown by the data in Table 3 and Figure 3. The ,"".ilrc1ation coefficient is exactly one and the sample monthly means of both money and gold stock changes are 2.6 million paper pesos for the prewar period and 4.4 million paper pesos for the 1927-1929 period. Thus, there was long-lived respect for the currency rule, and inflows and outflows of gold were fully tolerated and directly converted into changes in the money base.

More remarkably, there was even strict adherence to the money-gold rule in the float from 1914 to 1927, during a period of suspension of convertibility. In this period, as many core countries witnessed inflations and hyperinflations during and after W orld War One, Argentina maintained a key element of orthodoxy. There was no wild recourse to money printing, even if the exchange rate had drifted away from its anchor. Indeed, such drift in the exchange rate was unavoidable as most other countries had abandoned their pegs toa after 1914 and until the mid-1920s. Unilaterally, however, Argentina did what she could to stick to orthodoxy. As was analyzed elsewhere, the suspension of convertibility was foreseen in 1914 as a temporary political-economy decision to overcome the disruption of intemational trade and finance during World War One. This adherence to convertibility was all the more remarkable gi ven the enormous economic contraction in the years 1914-19 already noted above. All the same, changes in money tracked changes in gold for the entire period from 1900 to 1931.

13

.. ,' ,

...

•

Figure 3

Changes in Gold Stocks and the Money Base, 1900-1935

Gold Standard 1900:2-1914:7

40 ｾ＠ DM

.r

/

/oc

- - - ｾ＠ t '

--40 40

.,.

+ ＭＴＰｾ＠

Gold Standard 1927:12-1929:12

50 ｾ＠ DM

+ .ft

ç

ＭＭＮＮＮｾＫＭＭＭＭＭ

-60

-50 +-tI" 50 -100

+

-50 ｾ＠

Notes and Sources: See Table 3.

Float 1914:8-1919:12

-60

Float 1930:1-1931:3

100 - DM

_,t

ｾ＠

++

100

--100

4-DG

100 -100

:t

Float 1920: 1-1927: 11

100 - DM

4-T

+

t

---f

- -

-4- 100

100

-Float 1931:4-1935:4

(Caja rediscounts)

100- DM

+

T

++

ＺＺＴＭｾ＠

4-T

4--100 ｾ＠

After 1914 convertibility was suspended and exchange controls applied. N onetheless, there

were occasional outflows of gold for official purposes, and at the Conversion Office these were still accompanied by a strict application of the gold-for-peso rule. \O But for the most part, in this period 1914 to 1927 the Conversion Office worked in an asymmetric fashion: the monetary base augmented automatically when gold reserves increased but gold extractions were rarely allowed. Only in five months during 1914:8 to 1919:12 were outflows registered, and only for eight months during 1920:1 to 1927: 11 (see again Table 3 and Figure 3). Consequently, in periods of demand pressure in the foreign exchange market, notably in the recession of 1920-21, whilst adherence to the "rules of the game" would have allowed a large gold drain and a monetary contraction, in contrast the quantity of base money remained unchanged and adjustment instead took the form of a depreciation of the exchange

10 But here note that outtlows were very Iimited: 5 out of 65 months in 1914-20, and 8 out of 95 months in

1920-27. Thus the coefficients for the outtlows sample yield large standard errors. StiJl, the hypothesis of a unit coefficient cannot be rejected.

14

"

-? ' . . . f • '

,

•

rate, or, an increase in the paper currency premium relative to the paper-gold par (Figure 4 ).11

However, it was recognized by policymakers and even by the market players that resumption at par was the only possible steady state solution to the turbulence in the monetary and financiai markets. How could this tension be resolved?

A "one-way" gold standard, as this has been called, was still consistent with the long run goal of resumption at par. for a simple reason. During the boom years of 1900-14, Argentina had built up a huge gold backing for the currency; and from 1914 to the mid-1920s Argentina had husbanded that stock carefully via the "one-way" devices of the Conversion Office. By the mid-1920s,

resumption stiU looked credib1e because this strategy left Argentina with much stronger backing for the currency, at least as compared to many core countries, as the gold-exchange standard was built. Thus, actions and beliefs reinforced each other during this "one-way" gold-standard regime. Actions were geared to the monetization of gold inflows to expand the quantity of money but officials at the Conversion Office were strict about targeting the gold-backing of the quantity of money to a minimum of 78% in the turbu1ent 1920-25 years. At the same time, core gold standard countries such as Italy, Netherlands, Norway, or the United States never surpassed a gold-cover ratio of 50% of the money base.

It is relevant to ask what price Argentina paid in terms of deviation from parity in the foreign exchange market given this curious "one-way" gold standard policy. Did the exchange rate rise weII above par, requiring a big appreciation to permit resumption? Not at alI. The most dramatic foreign exchange crisis in 1920-21 combined sudden declines in export volumes and the terms of trade. Even so, the premium over par in the foreign exchange market never went above 33%,12 European countries could achieve such close convergence to prewar parity only by the mid-1920s, when they started to resume the gold standard.

Several observers have noted that the European experience with floating exchange rates after 1914 were dismal in the absence of a well-understood monetary "straight jacket" to limit

expansionary monetary policies, with monetization ofte!': ｣ｾｲｩＢｩｮｧ＠ frem unsustainable fiscal gaps. 13

Conditions in Argentina could not have been more different. More orthodox than the core itself, this peripheral country, from 1914 through to resumption in 1927, lived by the rhetoric, and also the actions, of a govemment intent on the idea of resumption at parity. To this end, and notwithstanding the suspension of convertibility and the "one-way" gold standard, the regime in place was still essentially a metallic monetary regime, and the prevailing mentalité allowed no room for money

Ii See della Paolera (1994, 34-36).

I: In the opposite circumstance, of gold inflow pressure, Argentina was even more orthodox than other

countries; in the 1918-19 postwar years, the paper peso strengthened well above par before the Conversion Office

decided to permit and monetize incipient gold inflows. In 1918 and 1919, the paper peso in terms of the gold-peso

was quoted at 2.14 and 2.2, well below the 2.27 par value.

15

issues not fully backed by gold. Historical experience with profligate monetary policies in the 1880s

and before, and the inflation and economic chaos that ensued from such actions, lived on in the minds

of the Argentine policymakers. Fear of repeat inflation underpinned their strong faith in the religion of

the gold standard, which they saw as an effective way for the Argentine authorities to tie their own

hands. Even the severe 1914-19 economic downturn could not break this commitment.

Students who focus on the impact of World War One on the Argentine economy might dispute

the idea that fram 1900 to 1931 the country maintained a metallic regime because of the emergency

laws installed in August 1914 to overcome the severe financiaI crisis. The purpose of the 1914

rediscount law was twofold. First, it established that the state-owned Banco de la Nación could

rediscount commercial paper from other private commercial banks by using up liquid reserves (that is

by declining its vault-cash-to-deposit ratio. Second, it established that, with the consent of the

Executive Power, the Conversion Office could also effect emergency issues (that is, non-backed

increases in the quantity of money) as long as gold reserves at the Conversion Office never fell below

40% of the outstanding money base.

The first measure was intended to delegate to the Banco de la Nación, a quasi-public bank,

and the largest of all banks, the microeconomic responsibility to forestall liquidity prablems in the

financiaI system. In that sense, one might say that the Banco de la Nación was a forerunner of the

Central Bank because it had a rediscount winàow. However, the action of rediscounting commercial

paper through this first provision in the law could only effect a change in inside money, or banking

money (M3), and not in the monetary base (MO) which was contralled by the Conversion Office.

More important, therefore, is the second provision of the law that allowed the Conversion

Office to rediscount commercial and govemment paper so long as gold backing stayed above a 40%

lower bound. Here we see, as early as 1914, and just 15 years after the 1899 convertibility law, that

there was a clear innovation: the design of an institutional capability that would permit the

Conversion Office to de-link gold and currency movements. Was it used at all? In the period to April

1931 the ｡ｩｾＺｷｾｲ＠ i:: ｡ｬｲｲＮ｣ｳｾ＠ ＺＭＺｾＢ･ｲＮ＠ With the exceptio'1 of three months in the economic crisis of 1925,

the monetary authorities never issued fiduciary notes.14

13 See the fascinating articIe by Eichengreen and Temin (1997) that analyzes the rhetoric of policy debates over

the resumption of the gold standard in core European and North American countries in the 1920s.

14 The conservatism of the Conversion Office contrasts markedly with the more activist behavior of the Banco

de la Nación. The bank rediscounted commercial paper in a countercycIical fashion (Salama, 1997). However at its maximum, rediscounts never represented more than 4% of a broad definition of money such as M3.

16

QIIUOTECA

MARIO HENRIQUE _ _

Figure 4

Exchange Rate, Pape r Pesos per Gold Peso versus Par, 1900-1940

6.0 .

5.5

-5.0 .

4.5

-4.0·

3.5

3.0

-2.5 .

2.0 .

1.5

-ｉｾｏｏＺｃｬ＠ 1905:01 l'1lU:\Jl 1'115:01 1920:Ul 192.:5:01 1930:01

Notes: Par is 2.27 paper pesos per gold peso until April 1935, and 4.96 from May 1935. Source: Baiocco (1937).

1935:01 1940:01

Let us examine the evolution of the money base and the actions of the monetary authorities. Figure 5 shows the evolution of the money base (MO) and its components during the years of the Conversion Office (January 1900 to April 1935) and the first years of the Central Bank (May 1935 to Oecember 1940). The constancy of the fiduciary issue of the Conversion Office is immediately apparent in the first component of the money base, and, above it we see the gold-backed component:

the strict gold-money rule applied. Except for a small blip in 1925, and a few other months (where lags kept gold and money slightly out of synchronization), there were no new money issue not covered by gold at the Conversion Office. 15 For this reason we found the very strict correlations between gold and

money seen in Table 3 and Figure 3. Consequently, we think of there being no significant change of monetary regime until 1931.

The old regime survived fram 1900, thraugh the years ofturmoil from 1914 until resumption in 1929, and even beyond the last departure fram convertibility in 1929 until 1931. But the previously

17

. '," _ _ ｾ＠

noted deflationary pressures and fiscal strains unleashed by the World Depression were sufficient to expose the suffocating potential of the gold standard "straight jacket." Moreover, given the

circumstances, mere suspension of the gold standard did not suffice to provide release from gold standard mentalité. If gold had to be spent in a fiscal rearguard action, the traditional operations of the Conversion Office would have led to major adverse monetary shock, unless the rules of the game changed. They did:

Maintenance of liquidity was not simply a matter of ending convertibility. On the one hand, even after the abandonment of the gold standard, some countries such as Argentina shipped gold abroad to service the externai debt and sold foreign exchange to stem the currency depreciation. On the other hand, as early as 1931 South American monetary authorities began to adopt measures which Professor E. W. Kemmerer and Sir Otto Niemeyer would have found unsound. Thus, the Argentine Caja de Conversión, whose old and only duty was to exchange gold for domestic currency and vice versa, began in 1931 to issue domestic currency in

exchange for private commercial paper. By 1931 the old Caja even issued domestic currency against treasury paper (Díaz Alejandro, 1983, 16-17)

It is therefore very important to note that the association between money and gold totally breaks down for the period after April1931 (Table 3 and Figure 3). Gold inflows ceased at this point, but subsequent gold outflows for fiscal use were sterilized. In many months there we no gold

movements, but the Conversion Office unilaterally changed the nominal stock of money. Accordingly, we conslder this the áeclsive regime change for Argentine economic policy, certainly during ｬｦｬｴｾ＠ Great Depression and possibly for the twentieth century as a whole. It was then that rediscounts began at the Conversion Office (Figure 5). These rediscounts offset the gold losses that began in 1929, and the fiscal use of gold in the subsequent years. The rediscounts began under the auspices of the 1914 law, and were soon augmented by a new series of rediscounts of treasury paper under the "Patriotic Loan" legislation of 1932, which further enhanced the power of the Conversion Office to make new fiduciary issues via open market operations.

Thus it was in April 1931 that the macroeconomic and, specifically, the monetary regime shifted from being a metallic regime to a fiduciary regime. At the end of that year, after just nine months, fiat notes rediscounted by the Conversion Office went from zero to represent a 30% of the outstanding monetary base. In 1932, the stock of fiat issues rediscounted by the virtue of the Law of 1914 accounted for 35% of the monetary base. At that time, the gold backing had already fallen from the targeted 78% in 1929 to 43%, a ratio slightly above the lower bound limit allowed by the law. In this context, it is instructive to examine expectations as indicated by behavior in foreign exchange markets: in 1930, the gold premium stood at a maximum of 28 % still below the 1923 mark. However,

15 Note that during World War One, and to a lesser extent in the early 1930s, difficulties in shipping gold out

of Europe meant that incipient gold inflows to Argentina were held in European vaults (Legaciónes), but were the

property of the Conversion Office. These were counted as gold backing, and are denoted separately in Figure 5.

18

" , ' . ' . , '

'.,': -:.' ｾＭＮ＠ ｾ［Ｚ＠

Figure 5

Composition of the Money Base, 1900-1940

2.000

1.800

1.600

1,400

-[J Gold+Reserves (BC)

IJ Fiduciary Issue (BC)

[J Gold (Legaciones)

[J Gold (Caja)

• Redisc (PL) • Redisc (Caja)

• Fiduciary Issue (Caja) 1.200

1.000

800

600

400

-200

o

1900:01 200

-1905:01 1910:01

Note: Units are millions of paper pesos. Source: della Paolera and Taylor (1998).

1915:01 1920:01 1925:01 1930:01 1935:01 1940:01

when the authorities sterilized the fiscal outflow of gold in 1931 and began rediscounting, the premium skyracketed to 819c and never reverted to a 10wer value (see Figure 4).

The reaction of the foreign exchange market was a (very rapid) manifestation of beliefs that had changed: age:i.t:; ＺｾＨ［［ＺＭ･｡ｳｩｮｧＡｹ＠ were comi:1g to see the delinking of gold and currency as an increasingly permanent phenomenon. In a moment, when we will examine the mechanics of institutional change and the impact of monetary policy, we will see that the macroeconomic regime saw a drastic quantifiable change in 1931. We argue that this precocious heterodox approach by policymakers might have helped Argentina avoid a severe economic collapse in the Great Depression of the kind seen in so many other countries. The decisive delinking of gold flows and the change in the quantity of money are evident fram the figures presented in the fourth quadrant in Figure 3. For the

1931-1935 period the average annual change in the gold stock was -6.7 million paper pesos but the average annual million pesos change in the quantity of money was +0.6 million paper pesos.

However, before pursuing a quantitative analysis we think it important to note exactly how remarkable it was that policymakers were able to make such a dramatic change in regime given the

decades of adherence to monetary orthodoxy. From a political economy point of view, how was it possible for such heterodox policies be contemplated in a country that had witnessed the Baring crash and had struggled for years to maintain credible monetary and financiai institutions?

The Political Economy of Reflation in Argentina: A

Deja- Vu

in 1929?

The very costly effects of deflation on the economy were nothing new in Argentine macroeconomic history. Was this just a seemingly unfortunate fact? The only benefit was what could be learned from the experience.

In the aftermath of the 1891 Baring crash, the paper peso, previously equivalent to one gold peso (and one gold dollar) by the virtue of the 1881 Convertibility Law, suffered a depreciation of 274%. In spite of the severe misalignment, in 1893, and in the middle of negotiations with

international creditors to settle the externai debt situation, Argentine monetary authorities assured international investors that convertibility would be resumed at anytime and at par. Subsequently, however, there arose a political-economy debate on whether a gold standard regime should be restored at the original 1881 parity, or if, instead, a conversion was to be pursued at the then prevailing market exchange rate, thus accommodating the ､･ｶ｡ｬｵ｡ｴｩｯｮＮｬｾ＠ Urban sectors and

commercial interests tavored a convertibílíty plan ftxed at par, while exporters and industrial sect0rs called for a higher nominal exchange rate because they believed that any further deflation of the economy would undermine the profitability of the real sector.

The most important political-economy argument denouncing the damaging effects of deflation originated with Silvio Gesell (1862-1930). In an article entitled "Monetary Anemia" written in 1898, Gesell noted the problem of the debt-deflation trap, anticipating Irving Fisher' s original work by almost thirty years: "If money gets more expensive, debts increase in exact proportion to the rise in the cost of money. Nominally nothing changes, but materially the debt load increases. With the prospect of having to pay triple what one received, who will dare go into debt to start a new industry in the country? ... The increase in the value of money is the common cause for ali the country' s economic troubles .... " 17 Gesell' s fear of deflation carne to be the dominant view, and the ide a of a

resumption at par was abandoned. Finally in 1899 Argentina regained the Gold Standard at a new par rate of 2.2727 paper pesos per gold peso, using the revaluation of gold as a mean to restore a fixed exchange rate regime.

The influence of Gesell' s ide as on the importance of changing negative expectations to escape a deflationary trap was recognized a long time ago by Irving Fisher and John Maynard Keynes. In his

16 This section follows c10sely della Paolera (1994, 28-29).

17 See Silvio Gesell (1909), and especially the articles "La pIe tora monetaria" (written in 1909) and "La

anemia monetaria" (written in 1898). Note the discussion on pages 20-23.

20

book Stamp Scrip (1933), Fisher acknowledges Gesell's innovative ideas on a fiduciary regime as follows: "Medicine owes much to untrained minds, or at least to minds untrained in medicine. Even Pasteur, though a trained scientist, was not a doctor; and the laryngoscope was perfected, some say actually invented, by a great singing mas ter, one Manuel Garcia, of Spain. Silvio Gesell, who died recently, was a German businessman and quasi-economist. He lived in Argentina and wrote some of his many papers in the Spanish language. In 1890, while in Argentina, he proposed essentially that particular substitute for money which now bids fair to sweep this country [the United States] under the name of Stamp Scrip. Gesell before he died, accumulated a considerable following abroad; but it took the tortures of a depression to bring about any practical efforts to make use of his Stamp Scrip idea. There is much in Gesell' s philosophy to which, as an economist, I cannot subscribe, especially his theory of interest; but Stamp Scrip, I believe, can, in the present emergency, be made at least as useful an invention as Manuel Garcia' s laryngoscope."IR

John Maynard Keynes went even further in recognizing Gesell's influence in the General

Theory (193x): "The great puzzle of Effective Demand with which Malthus had wrestled vanished

from economic literature. You will not find it mentioned even once in the whole works of Marshall, Edgeworth and Professor Pigou, from whose hands the c1assical theory has received its most mature imbodiment. It could only live on furtively, below the surface, in the underworlds of Karl Marx, Silvio Gesell or Major Douglas ... "; then, in his Chapter Notes on Mercantilism, the Usury Laws, Stamped Money and Theories of Under-consumption, Keynes remarked that "[i]t is convenient to mention at this point the strange, unduly neglected prophet Silvio Gesell whose work contains flashes of deep insight .... "; after this Keynes spent about a dozen pages explaining Gesell' s specific contribution to the theory of money and ｩｮｴ･ｲ･ｳｴＮＱｾ＠

Gesell was certainly influential in the 1899 decision to resume convertibility at a higher nominal exchange, and he gained much attention for his critique of a fixed nominal quantity of money as a sub-optimal monetary rule. Yet, interestingly enough, he is barely mentioned by the other main influential intellectual figure in om story, Raúl Prel:>isch.

Prebisch was not only the key policymaker and architect responsible for the creation of the Central Bank in 1935, but, most importantly, he was also a most influential and respected economist of the day, perhaps the only one who could challenge the prevailing mentalité, and conceive of a policy change to avoid the severe consequences of deflation in line with Gesell' s ideas. Already in 1921, he had written a brilliant artic1e on the problems of the Argentine currency showing that he understood the costly effects of deflation.21l _{During the de facto govemment of General Uriburu (1930-31), Raúl}

18 Fisher (1933, 17-18).

19 Keynes (1931, 32, 353-358, 371-379). 20 Prebisch (1922).

Prebisch was the Undersecretary of Finance; but more significantly he was the creative figure of

policymaking.

A criticai task for Prebisch was to persuade other policymakers to come around to his position. Among the various protagonists, one crucial person was Federico Pinedo, a politician and economist who at first viewed deviation from monetary orthodoxy with suspicion. In an interview years later, we find an illuminating dialogue between Prebisch and Mateo Magarifios where Prebisch explains how the change in monetary regime :::ame about:

Prebisch: "I am going to give you an idea of how Federico Pinedo was converted to the

idea of creating a Central Bank. As I have said, when the General Uriburu spoke about the convenience of studying the creation of a central bank, in the report I wrote, Pinedo, in a series of conferences, disputed the idea. And he did it in a harsh manner. At the time, I knew him very little. But during the world depression there was a situation, when I was the

U ndersecretary of the Treasury, a catastrophic situation. The banking system was on the verge of collapse and we decided-I had the ide a-to invoke an old rediscounting law that was never applied. The law was approved during World War One, in the first week of panic that we experienced, and it allowed the Conversion Office to rediscount banking paper. We made it operational .... we stated that the rediscount law was to be applied .... Then Pinedo, who was in the opposition to the Provisional Govemment, in spite of having been a

revolutionary, enrolled with the Partido Socialista Independiente .... He carne to see me, I have now a clear picture of that moment; he said: "Prebisch, what mistake are you going to make?"

ﾫＨｾｑｵ･＠ barbaridad van a hacer?») He was agitated ... so nervous that he did not want to sit

down .. .I explained to him the criticai ｳｩｴｬｬｾｴｩｯｮ＠ of the Banco de la Nación. The Banco de la Nación was the institution which administered the Clearing House. The money that the Banco de la Nación had in the vaults was below the cash at the Clearing House. That tells you about the gravity of the situation.

Q

Prebisch: This was in the year 1932-no, in the year '31, in the depth of the world

depression. I gave him a huge amount of confidential information ....

Q And the issue was starting the project to create the Central Bank?

Prebisch: No, no, no. It was putting in motion a rediscounting law to allow the Conversion

Office to rediscount. And Pinedo believed that we would provoke inflation. I explained to him for two hours. I did not hide any secret .... He asked me a few questions and he started to become more calm. After two hours standing up I said: "Ok, Doctor Pinedo, you now know how is the situation. What would you do if you were in my shoes in the cabinet of the Ministry of ｆｩｲＮ｡ｮ｣ｾ＿Ｂ＠ A.nd he had the loyalty of saying, and this is why I admired him so much: "the same thing that you are proposing". He said nothing more. His criticism of the govemment ceased. For the first time he understood in the dramatic crisis that engulfed the country and the financiaI system. And he started to support the measures of the govemment. .. He

convinced himself that there was no backtrack, that the Argentine monetary system based on the automatic exchange of gold for paper and paper for goId could not function. But this was in the year 1931. Then 1932 and 1933 elapsed and, when he was minister of Finance in 1934 he called me, and he asked me to draft the proposaI for the creation of the Central Bank .... 21

In short, led by Prebisch, a consensus was reached by the economic intelligentsia in Argentina in which they clearly understood that the time had come to abandon the rigid constraints imposed by a metallic monetary regime. As elsewhere, the intellectuaI battle to break the oId regime was not easy

21 Own translation from Mateo Magarinos, Dialogas con Raúl Prebisch (1991), pages 108-109 and 110.

22

because of entrenched fears of lax monetary policy in the minds of people like Pinedo: "The then heterodox South American monetary policies, which started around 1932, were in some ways a 'relapse' into past inflationary propensities, a past which was meant to be exorcised by the adoption of gold standard rules ... indeed, memories of wild inflation under inconvertible paper during the late nineteenth century, memories still fresh during 1929-1931, hampered and slowed down the adoption of more self-assured and expansionist monetary policies" (Díaz Alejandro 1983, 18). But the late nineteenth century also offered other lessons, and the change of regime built on the ideas of Gesell and others. and on public perceptions informed by another deflationary spiral, that suffered in the 1890s crisis. No one was to perceive this as a temporary political economy decision, but as a fundamental break from the past: but could it save the Argentine economy?

Institutional Change: The Impact of Monetary Policy

In the discussion so far we have found strong prima facie evidence that if any policy actions mattered for Argentina' s economic recovery from the Great Depression, it was most likely monetary policies and the change of regime that were central. In this section we attempt to quantify the impact of these policies on the evolution of the macroeconomy in the 1930s using standard econometric techniques.

We firSl al(empt to assess the lil1k ÍJetweel1 monetary variables and nominally-deTlominated prices, by examining the evolution of prices and exchange rates using a vector-error correction model subject to a long-run PPP constraint. We show that monetary policy, when finally used, did have the potential to effect transitory real devaluations to stimulate the home economy, a channel for recovery noted in Eichengreen and Sachs ( I 985) and Campa (1990). However, the impact of such policies was small in the Argentine case: counterfactuals show that even with an orthodox monetary policy the path of exchange rates and prices would have been little different. This follows from the fact that Argentine sterilization operations, although they offset some of the gold outflows, did not aggressi vely inflate the economy.

We next attempt to examine another channel, via the credit markets, focusing on the real interest rate effects of a change in monetary regime. This channel was studied by Romer (1992) for the United States, and we employ a similar technique here, using the Mishkin (1992) method for forecasting interest rates. Our results show a big change in 1931 in the path of real interest rates, and counterfactuals show that absent any such change, the persistence of deflationary expectations would have assured much higher real interest rates for several years, thus diminishing the chances of recovery. Thus the "Mundell effect"-to use Temin's (1989) terminology-was criticai in the Argentine recovery: discarding the gold standard convinced private agents that the risk of deflation was eliminated, and expectations changed accordingly. A concluding analysis of the real side of the

23